Pancreatic cancer is notoriously lethal. A full understanding of the oncogenes and tumor suppressor genes mutated in this cancer will aide both diagnostically and therapeutically. Genetic screens in diploid cells have been difficult until recently. Despite the existence of powerful selection systems, insertional mutagenesis of one allele can only inactivate the second copy by producing a dominant negative gene product, a relatively rare event. Most tumor suppressor genes however require bi-allelic inactivation. RNA interference (RNAi) is a novel technology that can destroy a given species of mRNA in cells. Libraries are now commercially available that target all known human genes, and are ideal tools for gene discovery if one has a powerful selection system, such as tumorigenesis. This proposal uses this technology to identify novel pancreatic cancer tumor suppressor genes. Hypothesis: Novel pancreas cancer tumor suppressor genes can be functionally identified using whole genome RNAi libraries. SA1: Construct cell lines. Specifically, transduce non-tumorigenic and weakly- tumorigenic pancreas cell lines with whole-genome RNAi libraries. SA2: Select tumorigenic cell clones. Using growth in soft agar and tumor growth in nude mice, and select cell clones that have become tumorigenic. SA3: Sequence RNAi's and test independent RNAi's. Identify the responsible RNAi by PCR and sequencing, and confirm suspect tumor suppressor genes using independent RNAi constructs. [unreadable] [unreadable] [unreadable]